Saturday, August 30, 2008

Response from author of "The ADVANTAGE Seeding Trial"

In response to my earlier post "Merck scientific debate hits bottom" I received the following from one of the authors of the VIOXX seeding trial paper The ADVANTAGE Seeding Trial: A Review Of Internal Documents, Ann Intern Med 2008", Dr. David Egilman, MD, MPH.

I posted Dr. Egilman's opinions below at his request. My own thoughts follow.

Dr. Egilman also posted numerous documents to and from Merck's lawyer at this link. (The tone of some of those documents towards Dr. Egilman is particularly disturbing, such as the one here.)


David Egilman, MD, MPH
To: Healthcare Renewal

I am one of the authors of the "Seeding" paper.

Merck executive
Dr. Jonathan M. Edelman, Executive director, Global Center for Scientific Affairs wrote:

There's no doubt that the ADVANTAGE clinical trial had a legitimate scientific purpose. That purpose was to assess the gastrointestinal tolerability of VIOXX compared to naproxen – a commonly used arthritis medicine with known tolerability problems – in the treatment of patients with osteoarthritis in a primary care setting, and for the first time allowed patients taking concomitant aspirin to participate.

Dr. Edelman's claim that ADVANTAGE was the first trial that permitted aspirin use with Vioxx is incorrect. Study 058, which was part of Merck's application for new drug approval submitted in November 1998, permitted the use of up to 325 mg of aspirin per day. ADVANTAGE only permitted up to 100 mg per day. Study 058 was a 30 week trial. ADVANTAGE was only a 12 week trial. Study 058 involved patients older than 80, comparing Vioxx to namebutone (another NSAID) and was completed in April 1998, about 10 months before ADVANTAGE enrolled a single patient. Study 078, the ghost-written Alzheimer’s trial also permitted aspirin, began in April 1998 and was scheduled to be completed by April 2000 before the end of the ADVANTAGE trial. More patients took low dose aspirin in the Vigor trial than the Advantage trial.

Further, ADVANTAGE was not designed to test the effect of aspirin on anything. Although aspirin was permitted, the examination of aspirin effects was not included as part of the study objective which Merck described: "[t]o assess the tolerability of rofecoxib (VIOXX) compared with naproxen for treatment of osteoarthritis." On the other hand, Merck Study 136 was designed to test the effects of concomitant aspirin and NSAID use. This study revealed that Vioxx offered no GI protection to low dose aspirin users who constituted the majority of potential Vioxx users.

By November 1998, when Merck sought FDA approval for an OA indication and three months before the ADVANTAGE trial enrolled any patients, Merck had already informed the FDA it had sufficient studies to prove Vioxx was efficacious for osteoarthritis patients. Merck's osteoarthritis trials (that it used to get FDA approval for use of VIOXX in osteoarthritis patients) had compared Vioxx to 3 other NSAIDS. The VIGOR trial, which tested Vioxx and naproxen in rheumatoid arthritis patients, started at the same time as ADAVANTAGE and used twice the dose of Vioxx than was used in ADVANTAGE. VIGOR was, therefore, a more valid scientific trial to determine relative safety of the two drugs.

This request for and subsequent approval meant that Merck could claim VIOXX was as effective for pain as naproxen. Beginning long before the Advantage trial Merck claimed Vioxx was safer than all NSIADS including naproxen.

While ADVANTAGE was the first study that compared Vioxx to naproxen in OA patients, Merck had completed trials comparing the two drugs in other pain patients.

It is for these reasons that the President of Merck Research Laboratories (MRL) Dr. Scolnick characterized ADVANTAGE as "intellectually redundant."

Merck wrote:

ADVANTAGE was a double-blind, randomized, controlled clinical trial with a legitimate scientific purpose designed to answer previously unanswered questions about the use of VIOXX in osteoarthritis in a primary care setting. It was not a seeding study.

Merck’s personnel called it a seeding trial internally but in an apparent effort to hide its true purpose wrote:

“I eliminated the reference to seeding. It may be a seeding study, but let’s not call it that in our internal documents.”

Merck apparently tried to make it look like a typical science-based clinical trial to mislead IRBs and journals. All seeding trials make some claim to some usually trivial scientific purpose. For example the question of whether or not Vioxx works better on people as a function of height is a scientific question. It is of course a trivial question. In this case Merck had studies that compared Vioxx to naproxen for pain control before the Advantage trial had started. Merck also had a larger trial that tested GI toxicity.

Merck wrote:

ADVANTAGE was important because although the earlier VIOXX clinical trial program provided extensive data on efficacy and safety, it did not include naproxen as a comparison Medication.

This is not correct. Merck had a study that compared Vioxx to naproxen for pain control before the Advantage trial started. This was presented to the FDA in the original New Drug Application and was published in October 1999.

Merck wrote:

The ADVANTAGE clinical trial was designed, conducted, analyzed, interpreted and published by the scientific department of Merck's U.S. Human Health (USHH) organization, Clinical Development (CDP), in conjunction with participating investigators. CDP was part of the Medical and Scientific Affairs department of USHH and was separate from the marketing department within USHH.

A more complete picture is:

USHH was the marketing division of Merck. It included sub-divisions including sales, marketing and a medical/scientific group called CDP. CDP performed “medical” studies that could be used by marketing to increase product sales. These studies were only conducted if favorable results could increase sales. They were not done to obtain new indications for drug use and Merck did not publish or provide the FDA with some of these studies (example 106).

Mr. David Anstice, a marketing executive with no apparent scientific training or background was the President of Merck's USHH division. In contrast Dr. Edward Scolnick, a physician/scientist of international renown, was the President of Merck’s research division which was called Merck Research Laboratories. Mr. Anstice the person ultimately responsible for the Advantage trial testified:

Q. I understand that. All I'm saying is, this is the study [Advantage trial] that was initiated and conducted by marketing, right?
A. Yes, by doctors in marketing.

A picture being worth a 1000 words I provide the Merck graphic for the organization of the Advantage trial:

Here is the power point description of marketing’s role in the Advantage trial:

While the marketing division of USHH set the study “objectives” CDP was in charge of the technical aspects of study implementation.

Further proof that ADVANTAGE served little scientific purpose is the fact that Merck had already claimed that Vioxx was safer than and as effective as all NSAIDS including naproxen prior to the initiation of the ADVANTAGE.

Merck did not rely on any ADVANTAGE results to make any claim about Vioxx efficacy or safety to the FDA. The lack of scientific purpose behind ADVANTAGE is demonstrated by the fact that Merck had withheld the ADVANTAGE results from its application to the FDA that requested a label change that would allow them to claim Vioxx was safer than naproxen. Merck only produced the results after the FDA made numerous oral and written requests for the safety data discovered during the trial.

Merck claims Advantage produced important scientific information:

In the end, ADVANTAGE showed a different gastrointestinal profile between VIOXX and naproxen that was unaffected by concomitant use of aspirin. This was an important medical result for physicians. In addition to measuring the GI tolerability of VIOXX, investigators also monitored patients for adverse events, which were required to be submitted to the FDA. Therefore, in ADVANTAGE, Merck was further evaluating any potential risks of VIOXX.

In fact the trial that Merck performed to test the effect of concomitant aspirin and Vioxx use (136) showed that aspirin use thwarted Vioxx’s GI advantage. (Gastroenterology. 2004 Aug;127(2):395-402.) The conclusion Merck reached based on the Advantage result was wrong precisely because Advantage was not designed to test this question. This conclusion was based on data mining.

Merck wrote:

We also want to underscore that the scientific purpose of ADVANTAGE was properly disclosed to physicians-investigators, participants, and institutional review boards, and Merck's business interests were clearly understood.

Merck never disclosed the marketing purpose of the trial which included the fact that Marketing “set objectives.”

Merck accuses the authors of "bias.”

In this open letter to the editors of The Annals of Internal Medicine, Merck would like to put in perspective the latest article by four authors who served as paid consultants to plaintiffs' lawyers in the VIOXX litigation against Merck. We are troubled by the biased article, which contains numerous inaccuracies, and wonder about the motivation behind this attack on Merck's scientific excellence and integrity.

Merck writes that the authors were “plaintiff witnesses” and “cherry picked” documents. As far as I know, all litigation ended about a year ago. None of the authors were paid on a contingency basis, the authors received no compensation from plaintiffs to work on the papers we have written and none of the authors are working on any Vioxx litigation at this time.

If money motivation equates to bias then it is Merck and its representatives who suffer from this problem. The same cannot be said of Merck or its representatives because while the authors have nothing to gain Merck’s future earnings depend, in some measure, on its reputation for honesty.

Merck does not explain how any theoretical bias impacted on the paper. They have not come forward with contradictory documents or testimony.

The authors agree that all the documents should be made public so readers can decide for themselves if we “cherry picked.” Merck, and Merck alone, continues to block the release of the documents.

Merck would not allow us to cite documents that showed that Advantage investigators were accused of fraud and at least one was convicted. Some of this data was deleted from the published paper but readers and editors were never informed of these problems.

It should be noted that eight Vioxx patients had heart attacks during the ADVANTAGE trial compared to one patient on naproxen. Merck never informed any of these patient volunteers of the cardiovascular risks of Vioxx before or after the trial although the informed consent stated that Merck would pay for treatment of drug related side effects. Unfortunately for some, this would have been funeral costs.

Merck wrote:

It is unfortunate that the authors and journal editors chose not to contact Merck before finalizing these publications. Had any of these individuals contacted Merck, factual errors could have been avoided.

I contacted Merck's main trial lawyer Ted Mayer and provided Merck an opportunity to comment on another paper related to the Advantage trial before it was submitted for publication. I did not send a pre-publication copy to plaintiff lawyers. Mr. Mayer provided comments and i included his comments in the final version.

Merck then complained about my contact to Mr. Mayer to Judge Higbee and requested that she sanction me for contacting their lawyer and then insisted that I make no further contact with them.

-- David Egilman


My own thoughts on all of this are as follows:

The lessons of the Annals article are clear. The scientific community and the public are simply fed up with cavalier practices in the biomedical industry. While the article may affect remaining VIOXX litigation as pointed out here, it also should have an necessary "cleansing effect" on drug R&D.

It was not clinicians and scientists outside the industry who earned it the public's animosity. It was, and continues to be, the industry itself. As Roy Poses observed, there cannot be evidence based medicine if the evidence is distorted by commercial zealotry. I think the public recognizes this.

Further, marketing personnel have no place in pre-approval, pre-marketing scientific studies such as clinical trials. If they have a scientific role, it is perhaps in post-marketing studies, especially post-marketing surveillance for adverse effects. The latter is a function the pharma industry performs poorly.

Even in that latter domain, there must be honesty and transparency, characteristics with which the biases of marketing and its leaders (e.g., towards maximizing sales) may interfere.

Merck's responses to this new Annals article seem the antithesis of science. Instead, the responses appear to be right out of the political and legal warfare playbook (where logical argument prevails - except when presenting logical fallacy is to the advantage of the protagonists) ...

The response has been this:

  • ignore the arguments and evidence [in this case, emails and documents from Merck itself],
  • issue an ad hominem attack,
  • issue slogans ("litigation in the name of science"),
  • present a Red Herring as a distraction,
  • issue accusations of bias,
  • issue accusations of cherry picking,
  • issue accusations of the authors taking evidence "out of context", etc.
and finally:

  • produce no context or new evidence of your own, hoping the above tactics will make that unnecessary.
All that's seemingly missing are ... puppets.

Merck seems to expect the "courtesy" of being allowed to be a participant in any issues written about its conduct, as if the authors of this paper are its employees who under employment policies must permit any pubs they produce to be reviewed internally for approval. I believe the company's executives forget that their authority ends at the front gate.

It is also an odd and even arrogant request, especially considering the way the authors' evidence (Merck's own emails and documents) was essentially ignored in the responses, and the insinuations promulgated against the authors as serving plaintiff's lawyers -- which also implies that scientific integrity and the interest of patients is not their motivation.

As I mentioned in my earlier post, this is classic ad hominem and is insulting to authors in my opinion, as well as insulting to any clinician or scientist. Why would anyone want to provide a company that acts in this manner the courtesy of pre-publication involvement in investigative journalism?

Another issue I find interesting is noted in the limitations section of the paper where it is stated that:

... we did not identify, despite an exhaustive and systematic search strategy, documents detailing discussions between the marketing and research divisions, nor could we identify documents describing contracts between Merck and collaborating companies.

As well as the Annals editor's obervations that:

... despite the large body of documents searched by the authors, they discovered few details about exactly how Merck's marketing division carried out ADVANTAGE.

This is surprising. From an information science standpoint, it would not be unreasonable to believe there might be more related to this trial of a major drug in emails or powerpoints or other documents. This document sparsity could mean that the documents did not exist, but it also raises a speculative question about the document corpus produced in Discovery.

Could "inconvenient documents" have selectively been eliminated from the corpus of documents before it was released, and therefore not be present in the corpus of documents the authors searched? Could the documents the authors did find represent ones that were "missed" by some weeding out effort?

Finally, I believe the entire tenor of Merck's responses - ignoring evidence, stating the trial was "published in a peer reviewed journal" as a Red Herring and a justification of its legitimacy while ignoring Sox & Rennie's (the Annals editors) accompanying commentary Seeding Trials: Just Say "No" , belitting the authors, etc. - is very disturbing.

From "Seeding Trials: Just Say No":

No one told Annals the true purpose of ADVANTAGE. We learned about it when we received a letter to the editor from Dr. David Egilman, who was a consultant to the plaintiffs' attorneys in the civil suits against Merck. He had access to publicly accessible trial documents, which included Merck employees' e-mail messages that disclosed the true intent of the ADVANTAGE trial ... The article provides clear evidence that the intent of ADVANTAGE was to increase prescriptions of Vioxx (the study outcome of greatest interest to Merck seems to have been Vioxx prescribing rates) ... The documents do tell us that deception is the key to a successful seeding trial.

When I review NIH grant applications, if I suspect deception I will mention it to the study section leaders, and I can assure that if deception is verified higher up the chain, that application will not get funded (at the very least). I can also assure that legitimate medical journals do not willingly publish studies involving deception.

Merck's responses to the Annals article seem to represent a belief that the public (laypeople and physicians) are stupid and gullible.

This is not science, and is not a path to a return to the corporate greatness and respect that Merck once enjoyed and had earned.

Finally, a Red Herring of my own:

-- SS

Thursday, August 28, 2008

Merger Mania Redux: the Case of the Carilion Health Care System

The Wall Street Journal published an article on how one not-for-profit hospital system came to dominate its market, and the effects of that domination on local health care.

How the System Became Dominant

The WSJ article documented how a hospital merger created a vertically-integrated health care system. Note that in the old days of "merger mania," there was a lot of buzz in the health care research and policy circles about how creating such integrated systems to benefit quality and access, and lower costs. This rationale appears below.

In 1989, the U.S. Department of Justice tried but failed to prevent a merger between nonprofit Carilion Health System and this former railroad town's other hospital. The merger, it warned in an unsuccessful antitrust lawsuit, would create a monopoly over medical care in the area.

After the 1989 merger, Carilion continued to operate Roanoke's two hospitals separately. It later consolidated the hospital boards and in 2006, transferred most of Roanoke Community Hospital's staff and services to a renovated and enlarged Roanoke Memorial Hospital.

The moves eliminated any hospital competition in Roanoke proper....

[Carilion CEO Dr Murphy] was convinced that the cost and quality of care in Roanoke could be improved if doctors worked in a more centralized system. In June 2006, he announced a seven-year, $100 million plan to transform Carilion into a multispecialty clinic, like the Mayo Clinic.

Carilion began approaching private physician groups, offering to buy their practices and pay their salaries.

We shall see what effects CEO Dr Murphy's advocacy of more centralization had.

Effects on Costs

The domination by a single vertically integrated health care system apparently lead to rising costs.

Nearly two decades later, the cost of health care in the Roanoke Valley -- a region in southwestern Virginia with a population of 300,000 -- is soaring. Health-insurance rates in Roanoke have gone from being the lowest in the state to the highest.

That's partly a reflection of Carilion's prices. Carilion charges $4,727 for a colonoscopy, four to 10 times what a local endoscopy center charges for the procedure. Carilion bills $1,606 for a neck CT scan, compared with the $675 charged by a local imaging center.

Alan Bayse, founder of a local benefits-consulting firm who has sold health insurance in the area for 30 years, says health-insurance rates in the Roanoke Valley used to be 20% lower than in Richmond, Virginia's capital, and the lowest in the state. Today, he says, they are the highest in the state and 25% higher than in Richmond, citing rate information from insurer Cigna Corp. Anthem, another health insurer, says its rates are 6% higher in Roanoke than in Richmond.

Mr. Lionberger, whose construction company has about 100 employees, says his health-care costs have risen 50% over the past three years, hampering his ability to compete with contractors from other parts of the state.

While the increasing domination by the system been associated with increased health care insurance rates, its leadership has aggressively pursued patients who failed to pay their share of its exaggerated bills.

The Roanoke City General District Court devotes one morning a week to cases filed by Carilion. In its fiscal year ended Sept. 30, Carilion says it sued 9,888 patients, garnished the wages of 5,478 people and placed liens on 3,920 homes. Carilion says the people it takes to court have the means to pay their bills.

When some patients don't pay their bills, Carilion places liens on their homes. Carilion says it doesn't track how many liens it has outstanding, but the close to 4,000 it filed in 2007 'is representative of a typical year,' Mr. Earnhart says. Carilion doesn't foreclose on homes and only collects when properties are sold, he says.

Dr. Murphy says Carilion only sues patients and places liens on their homes if it believes they have the ability to pay. 'If you're asking me if it's right in a right-and-wrong sense, it's not,' he says. But Carilion can't be blamed for the country's 'broken' health-care system, he says.

Ah, yes, the "broken" health care system made me do it, says CEO Dr Murphy.

Decreasing Access

One effect of the merger seems to be a squeeze on physicians who are not part of the integrated system, which presumably will decrease, not improve access to health care.

Some doctors who chose to remain independent say the number of patients referred to them by Carilion physicians plummeted. Carilion controls a large proportion of Roanoke's referrals because it employs a majority of doctors who make them, such as family practitioners, pediatricians and emergency physicians.

Joseph Alhadeff, an orthopedic surgeon who is a member of a private practice called Roanoke Orthopedic Center, says the number of joint replacements he performed dropped off sharply after he stopped getting such referrals from Carilion doctors, prompting him to plan to relocate to Pennsylvania. 'I spent seven years building up a practice and watched it evaporate in six months,' he says.

In fact, it appears that the integrated system was out to decrease the business of physicians not in the system.

Geoffrey Harter, an ear, nose and throat doctor at another Roanoke private practice, Jefferson Surgical Clinic, says Carilion-employed colleagues told him the hospital system asked them not to refer patients to doctors it didn't employ, calling such referrals 'leakage.' Keeping referrals within Carilion is lucrative for the hospital system because it ensures tests and procedures performed on patients take place at Carilion facilities.

Dr. Murphy says Carilion uses the term 'leakage' in internal marketing discussions and that he would rather see its doctors refer patients to other Carilion doctors to optimize their care. But he says Carilion doesn't require its doctors to keep referrals in-house even though it would be legal to do so.

Meanwhile, as the integrated system grew more powerful, its leadership appeared to use other tactics that have become all too familiar to readers of Health Care Renewal

Silencing Criticism

As tension between Carilion and Roanoke's independent doctors grew in 2006, a group of 200 doctors formed an organization called the Coalition for Responsible Healthcare to protest the Carilion Clinic plan. The group posted a petition on its Web site and put up billboards around Roanoke that read: 'Carilion Clinic. Big Dream. Big Questions.' The local newspaper, the Roanoke Times, covered the controversy in a series of articles written by its health-care reporter, Jeff Sturgeon.

A few months later, in March 2007, the Roanoke Times moved Mr. Sturgeon off the health-care beat after Carilion complained repeatedly about his coverage. Carilion says it communicated its displeasure to the paper's editors, but never asked that Mr. Sturgeon be reassigned. Carilion withdrew most of its advertising from the paper, but says it did that as part of a reallocation of its ad budget.

Follow the Money

As the system became more dominant, and more profitable, more money accrued to its top leaders, particularly, of course, its current and former CEOs.

In 2001, Dr. Murphy took the nonprofit hospital system's helm. Dr. Murphy, ... has a medical degree from Harvard but doesn't practice medicine....

Fueled by large, untaxed investment gains, Carilion's profits have risen over the past five years, reaching $107 million last year. Over the same period, the total annual compensation of its chief executive, Dr. Murphy, nearly tripled to $2.07 million. His predecessor, Thomas Robertson, received a lump-sum pension from Carilion of $7.4 million in 2003, on top of more than $2 million in previous pension payouts.

Carilion says Dr. Murphy's compensation is in line with comparable health-care organizations and notes he doesn't receive car allowances, a spousal allowance or club memberships. It says Mr. Robertson's pension accrued over a 32-year career at Carilion.

Members of Carilion's board of directors also seemingly profited from their relationship with the hospital system. The WSJ documented some major financial relationships among the hospital and companies in which board members had an interest.

A large part of the clinic conversion's costs have involved the construction of a new medical campus around Roanoke Memorial Hospital that began several years earlier.

The lead contractor building the site is Swedish construction giant Skanska. But one of the project's biggest beneficiaries has been J.M. Turner & Co., which is owned by Carilion board member Jay Turner. Carilion says it paid J.M. Turner a total of $14.9 million in direct contracting work from 2004 to 2007.

Dr. Murphy says Carilion's board authorized 'arm's length work' with J.M. Turner, but adds that "a case could be made that we shouldn't award work to J.M. Turner to avoid the appearance of impropriety."

Mr. Turner isn't the only Carilion board member with a financial stake in the new medical campus. Another board member, Warner Dalhouse, has invested in a hotel being built on the campus to accommodate patients and their families. HomeTown Bank, a local bank Mr. Dalhouse founded and of which he was until recently chairman, is financing the hotel's construction. Dr. Murphy and Mr. Turner sit on HomeTown Bank's board.

Carilion and Mr. Dalhouse say he didn't make his $130,000 investment in the hotel until after Carilion sold the parcel to Texas developers in early 2006. 'I wasn't dealing with Carilion. I was dealing with the new owners of that land who had paid fair market value for it,' Mr. Dalhouse says.

Management versus Mission

The Carilion Health System's statement of mission and vision are as follows

Mission Statement

Carilion Health System exists to improve the health of the communities it serves.


- Assure accessible, affordable, high quality healthcare that meets the needs of the community
- Motivate and educate individuals to improve their health
- Champion community initiatives to reduce health risk

Yet, the investigative reporting by the WSJ has shown how the increasing domination of local health care by the Carilion Health System has made health care less accessible and less affordable, despite its leaders' proclamations to the contrary. It has "educated" the local newspaper to reduce its critical coverage of the system's activities. Thus, the bigger and more dominant the health care organization, the more mission-hostile its leadership is likely to be. Meanwhile, increasingly mission-hostile leadership tends to become increasingly lucrative for its practitioners.

I submit that if we really want better quality, more accessible, more reasonably priced health care, we need to bust the new health care "trusts." We need smaller health care organizations with ethical leadership truly devoted to the health care mission. We need organizations whose governance is representative of key constituencies; accountable to patients, health care professionals, and the public at large; open and transparent; and which upholds clear ethical principles.

BLOGSCAN - "Be the Power," Never Mind the Data

Available in 3 parts on the Clinical Psychology and Psychiatry Blog (here, here, and here), and in one part on PharmaLot, is a video apparently used for training Merck pharmaceutical sales representatives to sell Vioxx (rofecoxib) to doubtful physicians. The strategy seemed to be to use flash and bluster to avoid dealing with concerns that Vioxx was no more effective a pain reliever than generic non-steroidal anti-inflammatory medications (NSAIDs) like ibuprofen and that Vioxx caused adverse effects, including myocardial infarction. Although waving around graphically fancy sales brochures was advised, dealing with the shaky data for clinical research (some of which later turned out to be gimmicked in favor of Vioxx) was not. Tell me again that story that pharma sales reps serve an educational function.

Tuesday, August 26, 2008

No Denying the Haunting of the Medical Literature

There was a remarkable series of letters in the latest JAMA(1). They, in turn, were responses to an article by Ross et al about ghost and guest authorship of articles about rofecoxib [Vioxx, by Merck](2). (See our post about that article here.)

Two of the letters were notable because they came from the lead authors of articles that Ross et al described as ghost-written.

First was a letter by Hawkey, which said in part,

From my own publications, I was surprised to see references 79, 94, 96 and 108 as having been in a Merck publication status report since these were reviews independently invited by Lancet, Balliere, Gastroenterology Clinics of North America, and Gut respectively, and involved no contact with Merck beyond usual request for all relevant sources for information.

Second was a letter by Ferris, Galasko, and Kirby, which said in part, (referring to an article by Thal et al discussed by Ross et al)

The 3 alleged guests in fact played substantial roles: conceiving and designing the study (Thal), confirming accuracy of the primary study end point as a member of the end point adjudication committee (Ferris), and enrolling 109 patient (Kirby). The paper was initially drafted by Merck coauthors in August 2003, after discussion of the results had previously taken place with Dr Thal and Dr Ferris in June and July of 2003. Drs Thal, Ferris, and Kirby were formally approached about being coauthors on the paper and an associated abstract in September 2003. Each critically reviewed the complete statistical report and contributed to the revising the final manuscript.

Before I go further, let me say a little about how scholarly articles are customarily written when no guest and ghost authors are involved. For all the articles I have authored, the first author was the researcher who was in charge of the overall study, or at least was substantially in charge of the study component reported by the article. The first author always wrote the first draft of the manuscript, possibly getting help from other authors or medical writers, but always being overall in charge. The first author always took responsibility for the final draft and published version of the manuscript. The second and third authors were generally those most involved with the relevant project, and with rewriting the manuscript.

This process seems very different than that used to write the articles discussed above.

Note that neither Hawkey nor Ferris, Galasko and Kirby claimed to be the principal investigators of the relevant projects, to have written the first drafts of the manuscripts, or to be those with the final say on the final drafts or published versions of the manuscript. Furthermore, none denied that Merck authors wrote the first drafts of the relevant manuscripts or that Merck authors had substantial responsibility for the content of the published papers. Hawkey claimed to have had "little contact" with Merck, but that did not preclude Merck authors from writing the first draft, or taking substantial responsibility for the content of the manuscript. Ferris, Galasko and Kirby claimed to have had some roles within the research project, but did not individually or collectively claim to be the project's principal investigator or to be substantially iin charge of the project, and did not deny that Merck authors wrote the first draft and took substantial responsibility for the content of the manuscript.

Thus, these letter writers did not deny that they had failed to do the work and take the responsibilities normally entailed by first authorship of scholarly manuscripts. Nor did they provide any evidence that the claims made by Ross et al about the ghost authorship of their papers were incorrect.

On the other hand, Hawkey and Ferris, Galasko and Kirby did take the opportunity to fulminate about the original JAMA article. Hawkey criticized Ross et al for failing to "systematically evaluate" all the articles in their Table. Ferris, Galasko and Kirby decried "this unsubstantiated allegation of guest authorship" which raised "questions about the flawed methodology in the study by Ross et al."

It is ironic that the financial disclosures for the letters by Hawkey, and by Ferris, Galasko and Kirby used 24 lines of type. Dr Ferris alone disclosed being a former or current consultant to 32 pharmaceutical, biotechnology or similar corporations.

What is most troubling, in my humble opinion, about the letters discussed above is that their authors could not bring themselves to directly address the core issue raised by Ross et al: evidence that they used their good names and academic reputations to cloak articles generated for commercial purposes in the guise of scholarship.

The larger questions are: how much of the clinical research database and the body of scholarship on which physicians and patients base their decisions was commercially generated, but deceptively promoted as produced by academics only interested in the discovery and dissemination of the truth? Furthermore, how many respected academics have built their careers and reputations on the work of unnamed employees of commercial health care firms, and thus what should we make of their supposed scholarship? And finally, on a personal note, how many honest academics have been crowded out of the competition for the support and dissemination of their work by faux scholars who claimed as theirs work done by others for commercial purposes?

As was said on the PharmGossip blog, it may be time for a truth and reconciliation committee that will allow us to put these deceptive ways behind us. To better care for patients and better advance science, we need medical scientists who put allegiance to the truth ahead of allegiance to their personal financial gain.

ADDENDUM (27 August, 2008) - See also comments on Science-Based Medicine blog.

ADDENDUM (2 September, 2008) - See also comments by Dr Howard Brody on the Hooked: Ethics, Medicine and Pharma blog.


1. Guest authorship, mortality reporting, and integrity in rofecoxib studies. JAMA 2008; 300: 900-906. [Link here.]

2. Ross JS, Hill KP, Egilman DS, Krumholz HM. Ghost authorship and ghostwriting in publications related to rofecoxib: a case study of industry documents from rofecoxib litigation. JAMA 2008; 299: 1800-1812. [Link

BLOGSCAN - Stanford's CME Money Goes Into a Pool

On the Carlat Psychiatry Blog, Dr Daniel Carlat analyzed the new Stanford policy that would direct all commercial funds supporting continuing medical education (CME) into a single pool. As Dr Carlat pointed out, the devil may be in the details of this policy, which still allows connection of funding to some (admittedly broad) particular clinical categories. Furthermore, although the policy would apparently sever direct connections between funding from particular companies and particular CME courses, I do not see how it could prevent informal communications between commercial sponsors and academics that could shape the subject matter and content of these courses. Such communication is now easy since so many medical faculty already have financial relationships with specific drug, device, biotechnology or other health care corporations.

Friday, August 22, 2008

Merger Mania Redux: Combination Would Lead to Windfalls for Blue Cross Executives

The Philadelphia Inquirer published a story about what seems to drive merger mania in health care. In Pennsylvania, the two largest health insurance companies in the state, both not-for-profit, Highmark and Independence Blue Cross, have been pushing to merge

Highmark Inc. and Independence Blue Cross would pay their top executives as much as $4.2 million more if they were allowed to merge.

Kenneth Melani, the chief executive of Highmark, who is expected to have the same job at the combined companies, would get a 31 percent raise, to $3.9 million from $2.97 million, including incentives, according to documents filed with the Pennsylvania Insurance Department.

Independence Blue Cross' CEO, Joseph Frick, who is slated for the role of chief operating office after the merger, would earn $2.94 million, the same as his current pay. In 2006, Frick's pay was $1.6 million.

The two biggest raises after Melani's $930,000 are planned for chief financial officer Nanette DeTurk and David M. O'Brien, executive vice president for government services, both from Pittsburgh-based Highmark.

DeTurk, who is in line to be CFO of the combined entity, could get a $635,098 increase in total pay, which includes possible annual and long-term incentives. O'Brien, head of Medicare operations, would get a $556,184 raise under the plan filed Thursday.

Lance Haver, director of Philadelphia's Office of Consumer Affairs, said yesterday that the raises were another 'indication that the Blues have lost their social mission and operate more like for-profit insurance companies.'

If the reason for the merger is 'to better serve the public, then they don't need to raise executive salaries like this,' he said.

The insurers have said that efficiencies from the merger, which would create the largest health insurer in state history and one of the largest in the nation, would spin off $1 billion in savings to benefit subscribers, the uninsured and other charities.

Part of the savings would come from the elimination of 745 to 1,200 jobs, the companies have projected.

An editorial in the same newspaper noted that while executives would enrich themselves, the benefits to others were not so obvious.

Short of the windfalls for the top execs, the proposed merger looks like a loser for other Pennsylvanians.

Employees: Up to 1,200 Blue Cross jobs are expected to be eliminated if the deal goes through. Mergers create job overlaps that lead to reductions and a big annual savings.

Of course, there's no overlap for the two chief executives, who instead will divvy up their duties and still get big paydays.

Customers: Independence Blue Cross and Highmark already have a stranglehold on their respective markets, which helps explain why they can jack up rates without much recourse by their clients.

The proposed merger won't lead to a reduction in premiums for customers, who have been socked with annual double-digit increases. The Blue Cross CEOs say the merger will have 'minimal' impact when it comes to reining in rate hikes.

Competitors: If the merger is approved, the prospect of any real competition among health-insurance providers in the state will disappear.

The merger would give the new firm a combined market share of roughly 65 percent - making it the largest private health insurer in Pennsylvania. The result will be a more dominant health-insurance provider, making it even harder for other existing insurers to expand while scaring any new competitors from entering the state.

The leaders of not-for-profit corporations are supposed to have a duty of obedience, the obligation to advance their organization's mission, and a duty of loyalty, the obligation to put the mission ahead of their self-interest when acting in their leadership capacity. Highmark proclaims its mission to be "to provide access to affordable, quality health care." Independence Blue Cross seems to have avoided committing itself to a mission, but by implication also should be about providing fairly priced, accessible health insurance. However, the main effect of this merger seemingly would be to enrich the two organizations' top executives, not to improve health care or make it more affordable or accessible. This example of merger mania reminds us that far too many health care leaders seem more interested in enriching and empowering themselves rather than improving health care.

We need health care leaders who actually care about health.

Making the governance of health care organizations, particularly not-for-profit organizations, more representative of key constituencies, more transparent, more accountable, and subject to clear ethical standards might encourage people who care about health to lead health care organizations, and discourage people who care mainly about their own wealth and power from doing so.

Health Care Executives Gone Wild: Financial Director of Shriners Hospital for Children Pleads Guilty to Mail Fraud

From the St Louis Post-Dispatch,

A former Shriners Hospital for Children executive this morning admitted stealing more than $800,000 that would have gone to the care of sick children.

Robert Brodzin, 41, was a certified public accountant and the director of fiscal services for Shriners - the hospital's main financial person here in St. Louis.

U.S. Attorney Catherine Hanaway called Brodzin 'the lowest of the low.'

'This case absolutely sickens me,' she said. 'While children were fighting for their lives ... he was stealing money.'

John Gillies, head of the FBI's St. Louis office, said Brodzin used some of the money for luxury cars like a Corvette, Hummers and a leased BMW, and to support his side business - two tanning salons.

Between October of 2000 and May of this year, Brodzin created three fake companies, then submitted 'payment requests' to Shriners. Because of his position, he was able to manipulate the internal accounting controls and hide the scheme – at least for a time.

In all, Brodzin reaped $828,973 from his scheme, according to testimony during Brodzin's plea hearing in federakl court in St. Louis this morning.

“You were really stealing this money... is that true?” U.S. District Judge Catherine Perry asked Brodzin. 'Yes it is, your honor,' he replied.

We seem to be having a run of hospital executives behaving badly in the US. Our most recent post on such a case was just three weeks ago.

Then, I proposed that it might make sense to subject leaders of health care organizations, such as hospitals, academic medical centers, pharmaceutical and device manufacturers, managed care companies, etc to a licensing process that could ensure they have some knowledge of health care, some first hand familiarity with health care on the ground, and at least an intellectual appreciation for health care's core values. Of course, I cannot promise that such a process would have kept someone like Mr Brodzin out of a health care leadership position. But maybe such a process would make at least a few snakes in suits uncomfortable enough to discourage them from a career in health care. Furthermore, maybe exposing a would be leader of a children's hospital to some of the sick children it serves would discourage him or her from manipulating the hospital's mission for private gain.

Wednesday, August 20, 2008

ADVANTAGE: Who was actually in charge of R&D? And other questions...

Roy Poses addresses many of the controversial issues raised by the Annals of Internal Medicine article "The ADVANTAGE seeding trial: a review of internal documents" in "Bad Seed: the ADVANTAGE Trial of Vioxx."

I wish to raise a few additional questions:

The Wall Street Journal and others pointed out that Edward Scolnick, MD, President of Merck Research Labs and largely seen as the #2 official in the company, worried that the ADVANTAGE study risked disclosing data to the Food and Drug Administration that could cause problems for Vioxx.

"Small marketing studies which are intellectually redundant are extremely dangerous," Scolnick said. [i.e., they pose a danger of non-meaningful adverse results due to chance - ed.]

I find this statement by Merck's head of R&D quite remarkable.

I believe Dr. Scolnick was, as I've written before at this HCRENEWAL post, the quintessential clinician/scientist of excellence. He apparently held the following titles:

(a) Merck’s Executive Vice President, Science and Technology, (b) President, Merck Research Laboratories (“MRL”), (c) member of Merck’s Management Committee, (d) member of MRL’s Research Management Committee, (e) member of the board of directors of Merck.

This raises the following questions:

  • If the President of the Research Labs, ostensibly the head of R&D at the company, could not stop such a "seeding trial" study of a drug not yet released and/or its publication, then who was actually in charge of R&D?
  • Was the go-ahead for the study approved by erstwhile CEO Gilmartin? The Board?
  • Was there internal conflict at high levels over this and similar issues of control?
  • Why did Roger Perlmutter, MD, PhD, former Merck Executive Vice President of Worldwide Basic and Preclinical Research -- who was believed to be Merck Research Labs (MRL) leadership heir apparent -- suddenly leave Merck in 2001? Could it have something to do with perceived lack of control of the MRL President's office over pre-release drug studies?
  • What other such "seeding studies" on other drugs were conducted, and who authorized them?

Several corollary questions also arise:

Merck rationed access to tools essential for R&D and new drug discovery to its R&D scientists on the order of several million dollars annually. Tools the science heads themselves wrote were essential to R&D were being denied them despite my best efforts as Director of the MRL Science Libraries to end the rationing through a comprehensive and (I thought) compelling gaps analysis. As an example of one site head's beliefs, a VP of Medicinal Chemistry wrote a note about substantially increasing desktop access to cheminformatics tools such as CrossFire and SciFinder as well as other specialty databases (as did other R&D leaders):

"These applications are invaluable to the productivity of medicinal chemists in my department. CrossFire, in particular, is highly useful as the primary source for chemical information. Most new Ph.D. chemists whom we recruit are accustomed to using these tools to facilitate their research, and in fact, it is embarrassing that I have eight new Ph.D.'s who have arrived beginning last August who still do not have desktop access to CrossFire or SciFinder. I cannot overemphasize the value of having these tools accessible to chemists at their desktop, and I would urge you to expedite their deployment as quickly as possible."

This raises the questions:

  • Why were "seeding studies" that the head of R&D himself called "intellectually redundant" and "extremely dangerous" being funded, while the needs of R&D and new drug discovery in particular were being underfunded?
  • The budgetary decisions for these tools, I believed, were ostensibly being made by the VP of Research Computing in MRL, an IT person lacking a science or biomedical background (an issue for another time). What pressures was that person being put under from other departments - e.g., those running "seeding studies" - to cut essential R&D budgets?
  • Was the budget decision not to end rationing of advanced informatics tools necessary for new drug discovery actually made by the VP of Research Computing? Or was the decision made by someone else to siphon money for other purposes, who had a different agenda?
  • Is Merck's pipeline thin due to such decision making? Did the company sacrifice its future for immediate gratification, not realizing the whole scenario would implode when the drug, VIOXX, was found to have "previously unknown" harmful side effects, an issue which itself has been the topic of much controversy and billions in litigation?

Related to VIOXX, other questions that arise relative to rationing of informatics tools are:

  • Could the rationing of tools essential to drug scientists in any way have been related to a desire by someone to hobble their ability to explore the issues related to drug side effects?
  • Why were many in the company who previously had live access to the webcasted, periodic R&D Research Council meetings suddenly cut off from access in late 2002 or early 2003? What, exactly, did the company fear from dissemination of such discussions about R&D?

Finally, the last questions that arise are:

  • Does all the above represent mismanagement?
  • Do the issues represent shareholder deception? (i.e., how can funding of "seeding trials" while simultaneously denying discovery scientist access to critical R&D tools be construed as the acts of a truly "research-driven" company? Indeed, Merck itself advertises itself to shareholders as a "global research-driven pharmaceutical company dedicated to putting patients first.")

Of course, the answers to all of these questions could be quite banal and ordinary. However, I merely report and raise questions. Perhaps someone needs to explore further, and decide.

-- SS

Tuesday, August 19, 2008

Bad Seed: the ADVANTAGE Trial of Vioxx

An article in the Annals of Internal Medicine by Hill et al(1), and the accompanying editorial by Sox and Rennie(2) have created quite a stir in the media and blogsphere. Even so, it seems worth reviewing the main points, and adding some comments.

Hill et al address the ADVANTAGE trial, published in the Annals in 2003.(3) This was a randomized controlled trial that compared rofecoxib (Vioxx, by Merck) to naproxen for patients with osteoarthritis. 600 investigators each enrolled a few (target 6) patients. The trial failed to show that rofecoxib had any advantage of naproxen in terms of pain relief (see our discussion here). Long after the trial report was published, it turned out that its first author had no involvement in the study until Merck presented him with a copy of the manuscript written by company authors describing it, making it one of the more prominent recent examples of ghost-writing (see post here).

Hill et al obtained access to numerous internal Merck documents that only came to light in response to discovery requests filed in litigation. They appear to show that the ADVANTAGE trial was actually a "seeding" trial, designed to market the drug by putting "its product in the hands of practicing physicians, hoping that the experience of treating patients with the study drug and a pleasant, even profitable interaction with the company will result in more loyal physicians who prescribe the drug." In support of that, they found these themes in the documents they examined.

- "The trial emerged from the marketing division with a marketing objective;"
- "Merck's marketing division collected, analyzed, and disseminated both the scientific and the marketing data; and"
- "Merck did not reveal the marketing purposes of the trial to participants, physician-investigators, and institutional review board members."

Remarkably, an internal Merck memo made these further points:

First, the trial was targeted to a select group of critical customers.

[These were] primary care physicians. The ADVANTAGE trial utilized this important group as investigators.

Second, the design of the trial focused on demonstrating the value of VIOXX to this important audience

So, the study was designed primarily not to answer a clinical or scientific question, but to target certain physicians as "customers."

Third, execution of the trial ... [involved] integration of the field, marketing, and CDP [Clinical Development Program].

Thus, the study was really run by marketers, not scientists of clinicians.


Finally, the results of the trial are being carefully tracked. An analysis performed at 6 months post launch demonstrated a significantly higher level of prescribing for VIOXX among primary care ADVANTAGE investigators compared to a control group of VIOXX 99 prescribers....

Thus, the primary subjects of the trial were really not the patients, but the "investigators." Really, this was a trial that showed that involving primary care physicians as "investigators" in a seeding trial caused them to prescribe the supposed study drug more often than physicians not involved in the trial.

Nonetheless, informed consent forms for the trial did not inform subjects of its underlying marketing objectives.

This article appears to be the first to provide evidence that pharmaceutical companies may deliberately disguise marketing efforts as clinical research. This is a real achievement, since obviously the companies involved make every effort to hide what they are doing, and it only through discovery during litigation did the facts come out.

The authors conclude that

At least 3 elements of seeding trials are harmful to science and society. First, full informed consent is not possible without disclosing the full purpose of the trial. Physicians and patients participating in ADVANTAGE were informed of the scientific objectives of the study, but the research protocol and informed consent templates indicate that they were not told about the key role of Merck's marketing division in the trial or the true purpose of the trial. Second, good research practice is at risk when the marketing division designs and conducts a study. In a seeding trial, in order to fulfill strong marketing objectives, the recruitment of several research sites with fewer patients per site may result in less quality control from investigators and use of sites that have less research experience than academic centers or community physicians' offices, which may be more accustomed to hosting clinical trials. Quality control and rigorous research conduct may not receive adequate attention when marketing is the primary purpose of the study. Third, the study may have little scientific merit. Around the same time as ADVANTAGE, Merck launched the VIGOR (Vioxx Gastrointestinal Outcomes Research) trial to be the definitive study of gastrointestinal toxicity. The FDA required Merck to conduct VIGOR before putting claims of improved gastrointestinal safety on the Vioxx label. Thus, the purpose of ADVANTAGE was neither to seek a new indication nor to perform postmarketing surveillance.


Failure to disclose the primary purpose of a trial has ethical ramifications for patients, physicians, and the design of clinical trials. Seeding trials like ADVANTAGE, in which the study medication has yet to receive FDA approval, may cause patient injury for marketing purposes.

The primary marketing objectives of seeding trials are hidden from the public, the medical profession, and institutional review board members, preventing them from making a fully informed decision about the balance of benefits and harms to themselves and society.

The importance of the article by Hill et al, in my humble opinion, goes beyond its clear demonstration that seeding trials do exist. The article demonstrates how clinical research may be twisted into marketing. The article demonstrates how an ostensibly scientific endeavor may be based on deception, when science is supposed to be about the pursuit of truth. The article demonstrates how the ethos of the huckster has replaced that of the physician in large organizations once regarded as ethical.

Once again, this is a reminder that patients, physicians and policy makers must be extremely skeptical of clinical research sponsored by organizations which have something to gain if the research turns out a certain way. Furthermore, it is a reminder to physicians that generous offers from organizations selling products, services or ideologies do not arise from altruism.

As Sox and Rennie suggest, physicians should "just say no" to seeding trials. Maybe it is time for society to "just say no" to letting those with products to sell run experiments on humans to test them.

See also comments on Alison Bass' blog, Clinical Psychology and Psychiatry, GoozNews, PharmaLot, Retired Docs Thoughts, See interviews with authors of the Annals article on the Carlat Psychiatry Blog and PharmaLot.

ADDENDUM (25 August, 2008) - See also comments by Dr Howard Brody on the Hooked: Ethics, Medicine and Pharma blog.


1. Hill KP, Ross JS, Egilman DS, Krumholz HM. The ADVANTAGE seeding trial: a review of internal documents. Ann Intern Med 2008; 149:251-258. Link

2. Sox H, Rennie D. Seeding trials: just say "no." Ann Intern Med 2008; 149: 279-280. Link

3. Lisse JR, Perlman M, Johansson G, Shoemaker JR, Schechtman J, Skalky CS, et al. ADVANTAGE Study Group. Gastrointestinal tolerability and effectiveness of rofecoxib versus naproxen in the treatment of osteoarthritis: a randomized, controlled trial. Ann Intern Med 2003;139:539-46.
[Abstract/Free Full Text]

Monday, August 18, 2008

Guidelines in Whose Interest? - the Withdrawal of the Children's Medication Algorithm Project (CMAP) Guidelines

In 2006, we posted about striking allegations that guidelines used in the state of Texas and elsewhere for management of adult psychiatric patients, the Texas Medication Algorithm Project, (TMAP) were influenced by payments by pharmaceutical companies to the state, and to the people developing the guidelines.

Now the Dallas Morning News has reported that a parallel set of guidelines for the management of pediatric psychiatric patients has been put on hold amid more allegations that these too have been so influenced.

A state mental health plan naming the preferred psychiatric drugs for children has been quietly put on hold over fears drug companies may have given researchers consulting contracts, speakers fees or other perks to help get their products on the list.

The Children's Medication Algorithm Project, or CMAP, was supposed to determine which psychiatric drugs were most effective for children and in what order they should be tried at state-funded mental health centers. In April, high-ranking state health officials gave researchers the go-ahead to roll out the guidelines.

A month later, the officials delayed the protocol, after Texas Attorney General Greg Abbott's office objected to it.

At most, the suspension indicates that state investigators fear fraud has occurred. At the least, it reflects nationwide unease with potential conflicts of interest between leading medical researchers and the pharmaceutical firms that fund much of their work.

Publicly, officials say it's because the state is suing a pharmaceutical company alleged to have used false advertising and improper influence to get its drugs on Texas' now-mandatory adult protocol, the Texas Medication Algorithm Project.

Privately, individuals with knowledge of the case – who spoke only on condition of anonymity because of the pending litigation – say the attorney general's investigation of possible fraud in the adult protocol has spread to the children's version.

At least four of CMAP's key developers – all affiliated with the University of Texas system, and all of them published child psychiatry experts – have received research funding from drug companies, or have been consultants and speakers for several different pharmaceutical firms, according to their own published papers and financial disclosure forms filed with the university. Drugs made by some of these manufacturers appear in the children's drug protocol.

The doctors say there's no room for improper influence when their reputations are at stake. If the drugs weren't effective, they wouldn't endorse them – and the research they conducted to craft CMAP wouldn't have been published in prestigious medical journals.

Dr. Graham Emslie, a UT-Southwestern psychiatry expert, said he never once witnessed improper influence from drug companies while he helped conduct CMAP research.

The news article listed four CMAP researchers and their financial relationships with pharmaceutical companies.

DR. M. LYNN CRISMON: The CMAP project director who heads UT-Austin's College of Pharmacy has received research funding or consulting dollars from at least 10 different drug companies, according to his published studies, including Eli Lilly, Janssen, and Pfizer. He said he could not comment on CMAP or the lawsuit.

DR. GRAHAM EMSLIE: The UT-Southwestern Department of Psychiatry researcher has consulted for several different drug companies, including GlaxoSmithKline and Pfizer. He has received research grants from at least three drug companies, including Eli Lilly and Forest Laboratories. University financial disclosure forms, where these drug companies are listed, report income in broad ranges. They indicate he may have made up to $125,000 from drug companies since 2004. He said the CMAP protocol was about evidence-based medicine, 'not the [drug] the most recent representative told me about.'

DR. STEVEN PLISZKA: The UT Health Science Center in San Antonio scientist has received research funding from Cephalon and AstraZeneca and has served as a consultant and speaker for McNeil and Shire. University financial disclosure forms, where these drug companies are listed, indicate he has made at least $130,000 in drug company speakers fees and consulting contracts since 2002. Dr. Pliszka said he didn't know CMAP had been delayed until a reporter asked about it. 'For any physician, the bottom line is, does their patient get better,' he said.

DR. CARROLL HUGHES: The UT-Southwestern's Department of Psychiatry doctor has received research funding from GlaxoSmithKline. University financial disclosure forms also indicate he was once an ad-hoc consultant for BioBehavioral Diagnostics, which designs equipment to test for behavioral disorders, and was awarded shares of company stock. He declined to comment.

Here we go again. I would note that the article included examples of the now familiar protests by a academic physicians with substantial financial relationships with pharmaceutical or device companies that such relationships have no effect on their judgments or decisions. But the psychological evidence suggests that the influence of conflicts of interest may not act at the conscious level. Furthermore, as Joe Collier observed [Collier J. The price of independence. Br Med J 2006; 332: 1447-9], "people who have conflicts of interest often find giving clear advice (or opinions) particularly difficult."

Furthermore, there is increasing evidence, (see this recent post, for example), that pharmaceutical and device companies cultivate "key opinion leaders" primarily to help them with marketing their products, and that they may be selected for their willingness to endorse these products as much as for their experience or intellectual prowess. It is the KOLs who generally get the consulting fees and speakers' honoraria. (Some academics may actually be consulted for specific technical or clinical expertise. But note that in this article, as is often the case, the consultants and paid speakers did not explain what they were paid to do, other than to market the payers' products.)

As I just said, --- there is a growing and ever more pervasive web of financial ties among academic medicine and commercial firms that make drugs and devices and provide various health care services. The resulting conflicts of interest suggest patients, physicians and policy-makers should be ever more skeptical about seemingly authoritative medical research, and medical education, practice guidelines, and health care policy based on this research.To make better decisions for individual patients, and better policies for populations, people, patients, physicians and policy-makers at least need clearer and more detailed information about all conflicts that may affect the evidence and opinions that influence their decisions. But might it not be simpler and ultimately better for people, patients, physicians and policy-makers to eliminate some of these conflicts so that we can begin to believe that researchers and academics are paid only to take care of patients, and create and disseminate new knowledge?

See also this post on PharmaLot.

Saturday, August 16, 2008

More Questions About Conflicts of Interest and "Surviving Sepsis"

Two years ago, we posted about questions whether the "Surviving Sepsis" campaign was driven by marketing as well as science. Then, we discussed a commentary in the New England Journal of Medicine which asserted that the campaign was part of an effort by Eli Lilly & Co to market Xigris (recombinant human activated protein C, or rhAPC, also known as drotrecogin alfa [activated]) for the treatment of sepsis.(1)

The Wall Street Journal just published an article questioning another aspect of that campaign. In 2001 a clinical trial in the New England Journal of Medicine by Rivers et al reported favorable results for early goal-directed therapy in the treatment of sepsis.(2) This therapy incorporated continuous monitoring of oxygen in the blood using a special central venous catheter capable of measuring oxygen saturation. The WSJ article first raised concerns that some of the patients initially entered into the study were excluded after randomization.

The medical-journal article analyzed 263. But it also said that 288 patients were 'evaluated,' of whom 25 'were excluded or did not consent to participate.'

Dr. Rivers, in an April interview describing how he proceeded, said 'all 288 are randomized' -- that is, placed into either the standard-therapy group or the other group -- 'and then some are excluded because they're not considered severe septic shock.'

But once patients are randomly placed in a group, they must be included in a final analysis, according to a fundamental principle of medical research. Twenty-five were not, the medical-journal article makes clear.

Further muddying the waters, Dr. Rivers, despite having said all 288 patients were 'randomized," stated in a later email that 'there were absolutely no patients pulled out before the analysis.'

The result appears to be a lack of clarity about exactly how this seminal study was done.

Hospital statisticians who worked on it had concerns about the data and recalculated the results with all 288 patients, according to people familiar with the events. When they did, these people say, the death-rate gap wasn't statistically significant.

Statisticians were especially concerned when they noticed that a relatively high proportion of the other 25 -- those not included in the final analysis -- were either conventional-therapy patients who survived or patients on aggressive therapy who died, say the people familiar with the events.

In 2005, a medical resident at Henry Ford doing follow-up research complained to senior hospital officials about the data in Dr. Rivers's study. The hospital convened a panel of inquiry, which concluded the issues had been satisfactorily resolved, hospital officials say.

The panel did a 'thorough investigation' and 'unanimously supported the conclusions,' the hospital said, adding that Dr. Rivers is 'an outstanding clinical investigator whom we support.'

The hospital declined to make the panel's report available. The medical resident who complained, Michael Donnino, declined to discuss the issue of the other 25 patients not included in the final analysis.

Dr. Rivers said that 'randomization was valid,' that 'all patients enrolled in the [early goal-directed therapy] study were appropriately selected, randomized and examined statistically,' and that 'there are no scientific integrity issues in the study.'
I should note that ideally randomized controlled trials should account for all patients randomized, and attribute their results to the group in which they were randomized, even if they did not complete the treatment to which they were randomized. Eliminating patients after randomization could bias the results of analysis, especially if patients particularly likely to have good or bad outcomes were selectively eliminated from one group or the other. The WSJ article raises questions about the occurrence of such bias, but does not prove it occurred.

Further complicating things, the WSJ article suggested that the authors of the article also had undisclosed conflicts of interest.

At the time the study at Henry Ford Hospital was published, in late 2001, the hospital was the owner of patents on the catheter used in the new treatment. Dr. Rivers had transferred the rights to the institution.

The New England Journal of Medicine didn't mention the patents when publishing the study. According to its editor, its policy at the time was to ask only whether an investigator had financial ties to companies involved, and Dr. Rivers said, correctly, that he had none.

Since the publication, catheter maker Edwards Lifesciences has made various payments to Dr. Rivers and his hospital. Such arrangements aren't unusual in medicine, but their existence can raise doubts about findings.

Dr. Rivers and the hospital over the years have received at least $404,000 from Edwards, the Irvine, Calif., company says.

For instance, Edwards paid $150,000 between 2001 and 2003 to Henry Ford Hospital for what Edwards terms 'nonexclusive patent rights [and] ongoing research' into blood oxygen and the mechanism of shock. Dr. Rivers, in an interview, described this as 'money I've gotten from the catheter...all for research.'

In addition, Edwards said, it has paid Dr. Rivers $158,000 to make speeches about his research, $20,000 to reimburse him for patent legal fees, $36,000 in reimbursed expenses and $40,000 in consulting fees.

Edwards Lifesciences said medical evidence has 'demonstrated that early goal-directed therapy yields significant reductions in sepsis-related mortality, lower organ failure rates and/or reduced health-care costs, particularly due to reduced stays.' In recent articles he has written, Dr. Rivers has disclosed receiving support from Edwards.

Also, Edwards Lifesciences was, along with Eli Lilly, a major sponsor of the Surviving Sepsis campaign.

In 2002, critical-care doctors from several specialties joined together to recommend treatments, calling themselves the Surviving Sepsis Campaign. Leading it were R. Phillip Dellinger of Cooper University Hospital in Camden, N.J., and Mitchell M. Levy of Brown University Medical School. Both, plus four others in the group, received consulting fees, honoraria or research-grant support from catheter maker Edwards.

Edwards helped pay for overseas meetings of the group in 2002 and 2004, which cost $861,000 in all. Surviving Sepsis endorsed Dr. Rivers's protocol as well as a drug made by another funder of the conferences, Eli Lilly & Co. Edwards said such support is 'routinely given to advance science and treatments for the benefit of patients.'

Finally, Dr Rivers also has also recently disclosed that he has been receiving money from Eli Lilly, for example, in a 2006 article, "Dr. Rivers has done consultant work for Biosite, Inc., Chiron, the Lilly Corporation, and Edwards Lifesciences over the last year."(3)

So, for an interval summary, recent guidelines promulgated by the "Surviving Sepsis" campaign include fairly enthusiastic endorsements for use of Xigris, made by Eli Lilly & Co, and early goal directed therapy, which requires use of a specific catheter made by Edwards Lifesciences (and other companies). These guidelines were based on research sponsored by these two companies which the guidelines touted as more definitive than it now appears to be. Furthermore, there is a web of financial links, that now appears increasingly complex, tying together these two companies, some researchers involved in these studies, their institutions, and the guideline writing process. These apparent conflicts of interest have raised questions about the integrity of the research on which the guidelines were based, and bias influencing the writing and deployment of the guidelines.

So it is becoming more difficult to determine the extent that marketing considerations influenced the "Surviving Sepsis" campaign, and some of the medical research on which its guidelines were based. Thus, the issue of how best to manage sepsis, already a complex one, now becomes more complex and less clear.

The larger lessons, again, are that there is a growing and ever more pervasive web of financial ties among academic medicine and commercial firms that make drugs and devices and provide various health care services. The resulting conflicts of interest suggest patients, physicians and policy-makers should be ever more skeptical about seemingly authoritative medical research, and medical education, practice guidelines, and health care policy based on this research.

To make better decisions for individual patients, and better policies for populations, people, patients, physicians and policy-makers at least need clearer and more detailed information about all conflicts that may affect the evidence and opinions that influence their decisions. But might it not be simpler and ultimately better for people, patients, physicians and policy-makers to eliminate some of these conflicts so that we can begin to believe that researchers and academics are paid only to take care of patients, and create and disseminate new knowledge?


1. Eichacker PQ, Natanson C, Danner RL. Surviving sepsis - practice guidelines, marketing campaigns, and Eli Lilly. N Engl J Med 2006; 335: 1640-1642. Link here.

2. Rivers E, Nguyen B, Havstad S et al for the Early Goal-Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001; 345:1368-1377. Link here.

3. Rivers EP. Early goal-directed therapy in severe sepsis and septic shock: converting science to reality. Chest 2006;129:217-218. Link here.

Friday, August 15, 2008

After Controversy Over Tobacco Money Funding Medical School, University President Steps Down

We have posted about the controversies arising from recently revealed research agreements between Richmond, Virginia based Virginia Commonwealth University (VCU) and tobacco company Philip Morris. These were first publicly discussed in May in a New York Times article. As we posted here, the main issues were that the research agreements themselves were secret; the agreements apparently gave Philip Morris control over all publications arising from the research, since they defined all products of the work as proprietary information belonging to Philip Morris; and that research for hire on behalf of a tobacco company, given that tobacco products have known severe health risks and no health benefits, seems to go against the mission of a medical school and academic medical center. We also noted that the university administration's apparent lack of qualms about its relationship with Philip Morris might have been related to its president's role as a leader of a tobacco company. (He sits on the board of Universal Corp, a tobacco buyer, processor, and distributor.) We later observed how little attention this subject has gotten in Richmond's major news outlet. Then, we noted that some university leaders were willing to open a public dialogue about the issue, thus exhibiting more transparency under criticism than has been shown by many other organizations. Most recently, we discussed how the Dean of the university's Medical School had allegedly actively sought grant support from Philip Morris for a center for research on women's health.

Today, the Richmond Times-Dispatch reported that VCU President Eugene Trani will step down next year. The article hailed Trani for his efforts to enlarge and expand the University. It also noted, however,

Trani was so intent on unifying the university in a shared mission that he forbid the mention of the Medical College of Virginia, renaming it the VCU Health System.

Furthermore, although the Times Dispatch has not published much about the controversy over the relationship between VCU and Philip Morris (see post here), it did acknowledge

But collaboration between VCU and Philip Morris also has stirred controversy because of the tobacco company's apparent level of control over research it sponsors through the university.

'The events of the summer are unfortunate,' [Professor of Political Science and Public Administration Blue] Wooldridge said, 'and would be a stain on [Trani's] legacy that didn't need to be there.'

A former president of the faculty senate, Wooldridge said he admires much of what Trani has done -- from building the engineering school to forging 15 alliances with universities in other countries -- especially while maintaining his own scholarship as a historian.

However, too often Trani hasn't considered the full costs to the university of his ambitions or listened to the informed opinions of others, Wooldridge said.

'I don't think the faculty has had their voices heard as much as they should.'

I hope that the impending retirement of President Trani will lead the VCU administration to rethink its apparent infatuation with cultivating tobacco as a source of funding for the medical school. The medical school's mission of promoting the health of individual patients, and the health of the public through teaching and research is ill served by financial arrangements with companies whose products only promote ill health.

Thursday, August 14, 2008



Recently we looked at some financial aspects of the boundary between Stanford University and Corcept Therapeutics. Dr. Alan Schatzberg, the chair of Stanford’s psychiatry department, received at least $510,000 from Corcept between 2000 and 2007. Senator Grassley asked why NIH received no disclosure from Stanford of Dr. Schatzberg’s realized gain from selling Corcept stock while he was Principal Investigator on a related NIH grant. The Senator also remarked on Stanford’s lowball estimate to him of Dr. Schatzberg’s equity in Corcept, currently between $5 million and $6 million.

Dr. Schatzberg is not the only academic to benefit from Corcept. Dr. Charles Nemeroff, a member of Corcept’s scientific advisory board, also did well. He exercised options to buy 60,000 shares on joining the board in 1998. Dr. Nemeroff diligently promoted Corcept’s drug. Following the style documented for Dr. Schatzberg, Dr. Nemeroff emphasized weak positive trends in the data while suppressing inconvenient negative analyses. In the fall of 2002, Dr. Nemeroff referred to the Stanford-NIH trials as “impressive studies indicating that ... (mifepristone) very effective in the treatment of psychotic depression.” The claims “impressive” and “very effective” are indefensible, and may even be fraudulent. Dr. Nemeroff’s exaggerated promotion occurred while the company prepared for its IPO, but he did not disclose his financial stake. Right after the 6-month SEC-mandated lockup period expired, Dr. Nemeroff sold 20,000 shares for $137,500. His cost was $6.60. Help me, what is the right term for this behavior?

A second member of the team who did well is Dr. Joseph Belanoff, Corcept’s co-founder and CEO. Like Dr. Schatzberg, he received around 3 million shares in 1998. Dr. Belanoff began selling as soon as the lock-up period ended. Between November 2004 and November 2006, Dr. Belanoff sold approximately 10,000 Corcept shares each month, realizing over $1 million. His cost for all those shares was around $80. An on-line financial service listed 6 additional quarterly sales, each of 30,000 shares, which netted over $900,000. Dr. Belanoff also received an ample corporate salary. His documented salary increased from $310,500 plus a 10% bonus in 2003 to $1,643,760 in salary ($411,008), bonus ($102,752), and stock options ($1,130,000) in 2007. Not bad, for someone with modest academic credentials.

Stanford University itself has a financial stake. Following a licensing fee of $47,000 that was paid to Stanford in 1999 along with 30,000 shares of stock, Corcept has been obligated to make non-refundable royalty payments of $50,000 a year. That adds up to $497,000 through 2008. Additional payments up to $250,000 are due to Stanford when certain drug development milestones are reached. Stanford has stated that it long ago divested itself of the Corcept stock. However, an on-line information service lists Stanford as a major shareholder as recently as 31 March 2008, with over 47,000 shares in the name of the Board of Trustees. In a second agreement, Stanford receives a $20,000 licensing fee, 1000 shares of Corcept stock, $10,000 a year in non-refundable royalties, and potential milestone payments of $350,000. All of this is in advance of any marketing of products by Corcept. So, Stanford is not exactly a disinterested administrator of the academic-corporate boundary between Dr. Schatzberg, NIH, and Corcept.

We should also note some unusual events involving Corcept stock movements. On 15 August 2005 Corcept stock received a research broker upgrade. Heavy selling followed the upgrade (volume was 613,000 shares, 31 standard deviations above the mean for the previous 10 trading days). For non-statistician readers, 3 standard deviations would tell us that the spike is almost certainly not consistent with the historical fluctuations in volume. So, a spike of 31 standard deviations means any such probability is astronomically small. The stock price moved up 34% from the previous day but quickly sank again. That same day, Dr. Schatzberg sold 15,597 shares for $109,179, at an intra-day high of $7 per share. The cost basis of those shares to him was $5.15.

On Friday 23 September 2005, heavy selling of Corcept stock occurred (871,700 shares). This volume was 35 standard deviations above the mean of the previous 10 trading days. On the following Monday, Corcept announced that a widely publicized study of mifepristone for Alzheimer disease was being halted prematurely for lack of progress. The company also announced slower than expected enrollment in a Phase III study of psychotic depression, for which they projected delayed announcement of results. Over the next month the stock price declined 16%.

These episodes of concentrated stock movement represent coincidences that remain to be explained. Was the August 2005 upgrade and selling spike in which Dr. Schatzberg participated a case of “pump and dump”? Was the September 2005 selling spike ahead of bad news a case of insider trading? Who did all that selling, and how did they know?

The questions just keep coming.